Text

Task 103:determine Events within TMI Primary Sys Insofar as They Might Indicate Core Configuration
	ML19221A625
Person / Time
Site:	Crane
Issue date:	04/06/1979
From:	Kaufman N; INDUSTRY ADVISORY GROUP
To:	;
References
	OLS-790406, TASK 103, TASK-103, NUDOCS 7905230464
	Download: ML19221A625 (13)
	v • d • e

/

y

. hJ tOA' )

N. Kaufman i

4/6/79/s TMI CORE ASSESSMENT L

J B-E-/ o 3

(.

.r CBJECTIVE:

Determine the events within the TMI primary system insofar as they might indicate core configuration.

METHOD:

Core density changes were inferred from reduced radiation transmission to intermediate and source range detectors.

Reactor coolant system pressure, reactor inlet and outlet temperatures, and steam generator pressure and leels were obtained from recorder plots.

Operatcr actions were obtained frca computer events monitor.

System behavior was theorized by a process of hypothesis and test for data consistency.

When possible, simple calculations were performed relative to heat up, boiloff, and volume changes to confirm reasonableness of process understanding.

Time inferred from graphs and/or recorder charts.

Estimated +3 minute accuracy

{

for event specification.

RESULTS:

b The sequence of inferred events one hour af ter turbine trip is attached. This sequence states as fact many items that are conclusions and hypotheses that seem to fit the available data, and it should be used accordingly.

Where facts are unable to be tested, a (?) is shown.

Also attached is a diagram sumarizing care density behavior in the period from 70 to 220 minutes as deduced from ion chamber traces.

In these traces, changes in transmissivity are interpretted as fluid / core density changes.

Changes in slope are considered important as indication of phase changes or elevation change. Also included in the diagram are plant and prccess information considered relevant and calculational notes.

CONCLUSICNS:

1.

Significant core damage occurs in the one hour period from 6 a.m. to 7 a.m. -

Ion chamber signals suggest significant voiding at this time.,' Significant superheat occurs.

Relatively little inflow to lower plenum is'r.oted during early portion of period when significant heating of voided core occurs.

Spike in ion chambers at 146 minutes suggests structure change,

- h Radiation noted in containment. More than -enough ti.ne and heat exists to cause metal reactions if cooling restricted.,

166 066 v m2ao gg j=>

I Dil CORE ASSESSMENT Page 2 i

~

Q CONCLUSIONS (Cont.):

2.

Greater damage on B-leg side of core than A-leg.

Significant flew noted on A-leg throughout period between 6 a.m. and 71.m.

as signaled by preferential heatup of outlet RTO and more rapid te.mperature decrease on A-leg inlet RTO during latter part of period.

3.

Core o P greater than combined AP of pump and steam generator.

Whenever pressure is reduced following 7 a.m., cold HPI and seal water ficw frcm pump and pump discharge toward outlet temperature detector in the pump suction.

This conclusion could be erroneous if the A-Generator outlet tap were cooler than mixture of lower plenum and HPI water.

Additionally, magnitude of temperature surges indicate A-leg in highly vaporous state.

7 Roverse flow occurs in A-leg starting about 2 p.m. and lasting until 10 p.m.

During this period, the plant is at low (500 psi) pressure and blowing down with flood tanks providing cold water to Icwer plenum.

4 Additio'nal damace during period following 7 a.m. possible, but not certain.

Outlet temperatures at superheat or saturation and cold water from HPI and/or core flood tanks in inlet plenum suggests core flow between 7 a.m.

i.I and 2 p.m.

Ion detectors suggest some density change, but nothing approach-ing the response noted at 6 a.m.

Frcm 7 a.m..(204 minutes) on, HPI, core flood, A-Generator and relief valve suggest ample heat removal capability.

On the other hand, highdP and reverse flew suggests low flow given lack of pump.

5.

Metal / Water Reaction Rate might be bounded by Hydraulic Data F, rom 130 to 176 minutes, the primary system pressure rises following valve closure.

This rise might well include contribution from H2 associated with metal / water reaction.

During this pr ^ssure rise, the in core density decreased by a factor of 1.5 to 2.5.

'th available pressures and temperatures, it should be possible to estimate H, fraction in steam and thus, reaction fraction.

It might also be possible to bound the energy added in 5ddition to decay heat which caused pressure rise, thus scoping exothermic reaction.

=

%W e

t j 166 06'l

4/ t> / / 9

/

TMI SEQUENCE OF EVENTS AFTER THE FIRST HOUR FOLLOWING TlJRBINE TRIP O

73 min.

- B-loop flow stops, B-Generator pressure begins to decrease, primary pressure quasi-stable at 1100 psi, with relief valve lifted and choke flow, T out near saturation T inlet subccoled, HPI off (?), A-Genera tor level slowly dropping, B-Generator level increasing.

90 min.

- Increase in B-Generator level occurs.

90 - 100 min.

- Cold fluid on B-Generator causes depressurization of primary system with density changes in core region.

Density decreases nearly linearly for 8 minutes as presssure decreases.

Magnitude is 1/3 of subsequent density change.

Temperature at saturation in both inlet and outlet legs.

A-Cenerator pressure begins to fall coincident with flashing, suggesting reduced heat removai Jue to reduced fluid density.

100 nifn.

- A-loop flow stops.

A quench occurs frcm fall back of cold water from head or B-loop or possibly from core barrel check valve leakage, although source is uncertain.

Quench inferred from nearly step wise density increase.

B-Generator pressure drop stops. A-Generator level begins to rise and pressure drop continues, suggesting continued low density in primary side of A-Genera tor.

.~)

100 - 116 min. - Pressure continues decrease, water new in core reheats. Rough calculation estimates 12 minutes required to reheat two times core volume up to saturation (factor of 2 is allcwance for inflow).

Inflow indicated by inlet temperature decrease.

Core outlet at or near saturation.

116 min.

- Voids begin to form in core.

Voids (rather than water density change) marked by slope change in source range detector. Outlet begins to show superheat.

Pressure continues to fall.

A-Genera tor level stabilizes, indicating some heat rejecti n.

Rate of pressure decrease in A-Generator decreases.

116 - 124 min. - Upper half of core voids, superheated steam in outlet. Hot spot probably steam cooled or voided at 121 minutes.

Signifi-cant steam flow in A-leg and B-leg core inlet at 124 minutes.

Significant steam ficw in A-leg outlet, less in B-leg.

124 - 132 min. - Rest of core voids at 128 to 132 minutes. At 132 minutes, gate valve on pressurizer relief valve closed by opera;or.

Pressure begins to rise.

Inlet temperature indicates little, if any, flow frcrn A-leg to inlet, but scue A-leg supe -heat rate of at least 10{F/ min. low from B-leg to inlet.

132 - 150 min. - Core appears voided with scrne censity reduction as pressure rises.

Outlet plenum contains superheated steam.

Some flow from i'-leg to lower plenum, li ttle or no flow from A-leg. At 146 minutes, the neutron detectors indicate one or more changes in core structure possibis.

166 068

e THI SEQUENCE OF EVENTS AFTER THE Page 2 FIRST HOUR FOLLOWING TURBINE TRIP.

g.,

U!

150 min.

Indication of steam on both generators since little or no heat removal.

Lower flow rate in B-leg finally raises outlet temperature detector to superheat indication 30 minutes later than A-leg.

Containment radiation spike reported. Adiabatic calculation (characteristic of dry out) would permit metal /

water reaction in 6 min. at hot spot and 10-15 minutes for fuel rod of average heat.

Simple calculation would permit core volume and downccmar volume at core height to be boiled out at 152 minutes.

150 - 176 min. - Pressure continues to increase with exponential appearance from 660 psi at 130 minutes to 2200 at 176 minutes.

Outlet temperatures in both legs indicate superheated steam.

Increased flow at inlet indicated by decreasing temperatures, with greater flow from the A-leg.

Little or no cooling indicated at either genera tor. Apparent exponential rise suggests autocatalytic H generation.

Core density indicated by ion chambers decreased 9

Ibss than expected from change in pressure, suggesting H2 presence.

176'- 180 min. - Quench of core and collapse of voids indicated.

Pressure discharge from relief valve.

Lower plenum water forced into cold legs, with greater flow to A than B (A contains less water) and some cooling by the A-Generator.

Main steam isolation valve and turbine bypass valve on B-Generator closed at 180 minutes.

Outlet temperature indicates superheat.

Quench suspected to be from water or steam from A-le" 180 - 204 min. - Quench water / steam reheats and, in 7 to 11 minutes, voids again begin forming in the core, although not as extensively as before (about one-half).

Little flow (possible backflow) in the A-leg as indicated by hot water at inlet temperature detector. More flew on B-leg and toward inlet plenum. Outlet temperatures at saturation.

Relief block valve opened at 204 min.

204 - 450 m!n. - HPI turned on.

Pressure drops to 1200 and is maintained between 1200 and 1700 psi until 5.5 hours5.787037e-5 days 0.00139 hours 8.267196e-6 weeks 1.9025e-6 months , when the relief block valve was closed. When valve closed, pressure rose to 2100 psi and remained there through period.

Throughout period, outlet temperatures were near saturation and inlet temperature cenerally decreased as the result of HPI flow.

Periodic surges of lower plenum water move into cold legs whenever system pressure decreases, more so in the A-leg than B-leg, indicating greater vapor volume in A-leg. Core voids or significant density decrease noted at 223 minutes, associated with pressure surge and backflow from lower plenun to inlet legs.

Slow correction or collapse noted in this case.

s v -)

166 069

i

,/'

-r TMI SEQUENCE OF EVENTS AFTER THE Page 3 g

FIRST HOUR FOLLOWING TURGINE TRIP 7.5 - 10 hours1.157407e-4 days 0.00278 hours 1.653439e-5 weeks 3.805e-6 months - Relief block valve opened at 7.5 hours5.787037e-5 days 0.00139 hours 8.267196e-6 weeks 1.9025e-6 months and pressure drops to 500 psi in 1.5 hours5.787037e-5 days 0.00139 hours 8.267196e-6 weeks 1.9025e-6 months .

At 8.5 hours5.787037e-5 days 0.00139 hours 8.267196e-6 weeks 1.9025e-6 months , the core flood actuates.

Outlet temperatures show superheated steam. Generators cease to remove heat for 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months starting at 8.0 hours0 days 0 hours 0 weeks 0 months .

During this 2 hour2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months period, inlet temperature decreases and outlet temperature is at superheat condition.

Small change in core density occurs at 9.6 hours6.944444e-5 days 0.00167 hours 9.920635e-6 weeks 2.283e-6 months .

10 - 16 hours1.851852e-4 days 0.00444 hours 2.645503e-5 weeks 6.088e-6 months - Reverse flow in A-leg begins at 10.5 hours5.787037e-5 days 0.00139 hours 8.267196e-6 weeks 1.9025e-6 months .

B-leg may have tried to reverse at 14 hours1.62037e-4 days 0.00389 hours 2.314815e-5 weeks 5.327e-6 months .

Unexplained level change in B-Generator at ll.5. hours that appears to be from charging, although logs and data sheets contain no reference to such action.

Primary side contraction pulls icwer plenum water into B-leg.

Even though HPI flow is increased at 12.5 hours5.787037e-5 days 0.00139 hours 8.267196e-6 weeks 1.9025e-6 months ',

reverse flow continues in A-leg.

Relief valve again closed at 13.5 hours5.787037e-5 days 0.00139 hours 8.267196e-6 weeks 1.9025e-6 months and pressure rise begins.

Increasing pressure pushes HPI water toward A-leg generator and pushes lower plenum water into inlet of B-leg.

Pressure increase and its effects extend from 13.5 to 14.5 hours5.787037e-5 days 0.00139 hours 8.267196e-6 weeks 1.9025e-6 months . During the episode, cooling occurs in A-Generator.

Some density change noted in core as pressure increased, then a general density decrease i

occurs.

Suspect hot fluid from upper plenum pushed through Core.

.\\

At 16 hours1.851852e-4 days 0.00444 hours 2.645503e-5 weeks 6.088e-6 months , A-loop Primary pump turned on and significant events end.

4.

O

~.

166 070

~~

~

LND1CMrfO

~

AdDUCCD

.?

0dAl!oTV==*>

s 3

IL7, 0

3. g' i

A h

O kCel* ' F"Luts! $ rCi*

4 W'

~

u%y w f,'

ps creas m 4.E ve' a k s

% h*

B Hverstrit' vasos s'oxn E

O 1

s 3

3

(

0 t

u

~

A-l Cop FL6W $7dI*, '

~

O o

fr-Ge verstor' l eecI JM'2 q

g%

h QucvcH fWe.wgd to 64 klEMA

~

N

{

c g

~1 D

C h

NE r

vce: Fatrt AGma

-~'i-y

'h o'~

~

Y* TN 7

.R-' 't

~

3 4 Y

x.

u.

~ A llG-QuT T&t! Plc.cc o ng

~ R1/sef valve b bez c Q g

~

3-

..y 3 -.........

y 4

e 1

- B-1.u-Our reest' tctg.c c o M~~$

m n

'e.

,2 2m 2

2-scs. OkyI4 G-ov7~

bA l 6 ftAA flew A J t s..k1 B-Ou d fD ne I

APPC/1M Allid" VAAYS GL-419 f', O ut us M l~ A c es /1 Ar oJ*

n

- ~

~

6 i ~y 3l A isce namy vosoeo, E k2

+=.

O 8 Sowenfca JJo/atel H

1

~

s.N e

AJE a) Voso,1 w4ev

~

o.

N l

9 C eM A a.,c.ter- /c ACAf4 g

9.,.

y 5. p w

so ---

,B g,. N.;%,4 g 4

~

fy yhf 2

r e lre f H IV C W

u ga oy av d doses.<v-s U

t i g,9 *

'}

2*e-c.*. fee

/e e e

i t.',, E {

fkm. H PIC srnorto KV lllfo s.on.

%g 4

1 166 071

~

,,e c

g 9.;.

s, a:

-j. *

. h, 99)^\\'.;p> -

i' T'

sE

+..

n XJ

,1 3--X a;

'3 r_.

L:2 9

y p

2

-X r-J.

2 I

lilGH PRESSURE g

INJECT 10:1 2

.;, 6tx ' 1;."

A.

s v

t..

2 FL0lslETERS

)

3' RESISTAllCE k

PU@ ^d b

PUMP Q

.gS g

TEMPERATURE'.

~

i

,\\%

I d/

~

liiDICATOR

,74 i

~

g 11 LETDOWil E'"

PRESSURE 3- -X STEAll STEAM

5. CORE FLOODIUG.,

GENERATOR VESSEL GEfiERATOR N0ZZLE 6

PRESSURIZER

'8' C0ilflECTION

' N THREE MILE ISLAND - 2

q. -

~

w 8

(0 Z

=

+. - - - -. <.,,,

,'!Oy[,r -

I D d'sj:o~,

o s

... - } l -.

l y

k 8 L CCP F~RitJ S T~ci/*

U.

DD f

,,, cre:.:c.

ir, s e iri c r~

s r s

3 7

si c-e <. - e c e vcsos a:<..

o 4 d th h

MM b M

O 3

-> Q L 80 r

~

E f

/C-laty /~ /. C w.l* ; G y %

~ ' '

fk 39

.5

)

f-Ge.,*.o?

t-

% a ' ;* r'~;ll U!%

c Q

TC b4 **

Q UC u c H Pi exa -e a b Q 7

0

/

q T..

D i

0,

.i. A

~-

/ O to s 7'Ct'M PG siieJ

't W

- - -.A ~h t-y s

- i.

c t:

N

/

I

/

3

,4 J/

^

M-O h

,,T.

~

n

- li LfG-C UT 7!rtr* P.CC.G C C

'b

- 6 la e 4 i ?. i < < h la e x ek.. ;

y n

Lq

-t O

$>b)'

)

2-Ltc* Cs,;~ Tret* rf*GCD l

Q n

L%

3g 2-wa. Dry!"c-our Ass T* / c u S J /s.

1.5 Co. O N d XD r-h 0

\\

~.

fjppcin' Hisstr VAAVE 5 ^^

1

- ~ +1r r*'F rscs n y Voso e 0, cut ~cn /~Pors li s. coq

$nt5;3 A 460

^ ', 1 4M o

G Ge~ cock ucisted

'N 1

~

(N N

- AJ E s.1 Voro.$ w h e J O

?

C ess4 h we. s er-re L G4, T4 H

w n

p

/

m.3 e F. 6 o -- -

/

1

%l s

/

eN t

m"

.n u.-

khk 8

__. r e js e s ?o lv e cre ?

r. ~ 1 s ; ;

i

~

-a c(

.p;

,a,o,.

o v d d,reA.r;,-

o s

[

a, 7 g e * {

,,{.,,

p h..un.

r ~.: rir.,: r :

cf 6 c v e 6 f<- /4-ee L

q LL y

er : i;,

n T

166 073

/

%lbf/2 H. Levenson

{

fM - E-I o y n

Step I Go to Natural Circulation-with Pressurizer Solid A.

Float core flood tanks at B&W set pressure, but with tanks solid tanks outside containment. This is to

~

and pressure coming from N2 expedite conversion to Step II if and when deemed desirable.

B.

All valves from primary system to Auxiliary Building closed except for occasional use of core flood tank fill line to make up losses.

Step II Conversion to Benign Building supply from core flood tanks.

A.

Remove high pressure N2 B.

Open Pressurizer Vent valve. Leave open.

Pressurizer Sample Line should have been rigged in reentrant mode back to contain-ment so that if the vent block valve cannot be opened, the V

sample line is a fall back.

.I C.

Eventual steady state may result in occasional bubbles rising through flooded pressurizer.

if desired) can be'added 1.

Water (plus additives such as H 022 via core flood tank fill line.

D.

Eventual state may permit isolation of steam generator secondaries and use of hot drain coolers as heat sinks--needs to be looked at.

1 C1 e

166 074

}

[] '

og

_?l

\\

/

q f'

//

ll 84e I

2 mr h

(b)fA A

T lE1h hu STEAR h

4' g

N GENERATOR i

\\

le f

l M

M 19 p? ~

f WLL

!!f V

j@

5eed ff j.$

k' tu

,ey p e

ud T.

' ' {

' itg-

.u t_

f b'

t

{

,bh I

E(

79'-82 s

O}1.1

,, e,. '

,[y E'e 8'-0"

/g Mgfpy{ft d

REACTO i

~ ;

f 5"

ly p 6'

{f 42'- 8 COOLAN g-

~

p uGP 706

-447

%:1 g

8 s

2

)

k% 0

@, ~lcop, f i;r

@,b 1$Q wruye i

l;

- i YQ4

.D

(

-n(

sjp FV a

4 u

p,;l

?

he 31 ' -O" s

d I

q BJ ' -O" i.

q f'.

h 1

"i ga J7 a

,3 2g *. 3 -

w

~-

g.,s gnagj k

o o

l

]{'. y

1 j-S'-0" 9'-0" f

p 1

sxpx N

NgN -b% s k

REACTOR CCOLANT 175 TEM AR2.ANGEMENT - ELEVATK TMEEI. MILE E5 LAND MUC1.ZAR STATION 12GT 2 l

~

1660'75 FIGURE 5.1-5

i i.1l: ll l

l V

g ii

i,;

l')

I w

l l j

-l l-Ef.-._f'_1l,'

s.

l l

l t

s

.q-

{d s

M. ',' i, f !

l.

i l

N'

3e M

2 a

c

\\

3

~

/.

\\W.x.NNNN'm\\.%%

a.,..

l l

7

%sg j.._

v

' ga\\ u' w w m s w; cs y

/sp n

\\

N 7 -i-i

!: , 1.

bi

' k \\\\

p 3-p f. e..._,

- i,..,,.,......h \\'C Q W U 2 x' n c',-V

l..

c 1

I l

J

-.l

j!

.l a.

7..d _i___.'_'.._j..*

s g!

q g

c m,.

ppm.

e w

j.. l.. !. :._. 4 g

'g A\\NNNNNNNNN\\@d i

~

i i

ii si i I

o I- ',.

g'3 i

- - - - ~ ~ - - - <

i t

j N

i Li

!.l

'l.-..i...!

i

\\

r r

., b h f

a n

1 m

h f..

?-$1W.'. -

( d, g6 4 %,.9 '.' Y f L

76 xWss piWNiNW%WM e

~

3n

.v

y. a.

N. P. Palladino 4/8/79 - /

~

Under Milt Levenson's proposed cooling mode, the pressure in the PV will start at zero psig and then start to increase slowly as the gas bubble above the core grows until this pressure is high enough to drive water out of the pressurizer at the rate required to balance the rate of bubble formation.

When the bubble reaches the top of the core-flood nozzle of the PV, gas will begin to accumulate in the line to the accumulator.

This may interfere with trying to measure the water level in the PV via the core flood tank.

When the bubble reaches down to the top of the hot leg, gas begins to bubble up to the top of the candy cane.

I believe that some of the P.S. water in the steam generator will flow out through the cold leg to replace the gas from the bubble (above the care) that goes to the candy cane. The PV level rises slightly.

Again the pressure of the bubble will rise to uncover the top of the hot leg port of the PV and more gas wil'1 go to the top of the candy cane. This process will continue until the PV level and the pressurizer surge line level reaches point A in figure on Page 2.

(See below for sketch of the uncovering of top t

of hot leg port as discussed above.

Uh

9,/

There will be a period of time when the flow through the core and to the pressuri:cr will stop. This will occur while the gas is building up enough pressure to drive the water level in the pressurizer surge line frcxs point A to point B.

This requires that the pressure in the PV gets high enough to overccme the static head in the pressurizer above point 8 plus any AP across the vent valve or about 20 psig. (See figure on next page)

{

~

166 077

' " ~ ~ ~ ~

_ e p q q q g n,p t y n n e - m <-

,e

/

' ~ Y T yg,D;7:f,R'?.'D r".

^-

N. P. Pa11adino

)j [

,.. A. ;',9t?T;'

4/8/79

.g 9..~

m Ty ~..

{O p.

~ y:::

~

F.

e.

7 l

1

.q-.

. // 2. '

L~,'.

- 4 42

.M t '

r -

p-j c

8 o

./

)

i (y'

1 When the level in the surge line reaches point B, the gas will perculate up through the' pressurizer water.

4 f

j During this period of time, the core will continue to both give off gas and to heat up and then boil at this higher pressure. The boiling will help increase the PV pressure faster and speed up the change in level from points A to B.

P e

e O

166 078

ML19221A625

Text

Navigation menu

Search